Since data assimilation can combine observational data and model simulation data to integrate their benefits,it has become an important global environment change research method.During the development process of data assimilation algorithms,the particle filter is free from the constraints of linear models and Gaussian error distributions,thus becoming the focus in the study of data assimilation algorithms.In this study,literature on the application and study of particle filter in the field of data assimilation is collected,arranged and concluded.Then the advantages of particle filter and its application in data assimilation are summarized.The main problems and corresponding solutions when applying particle filter in data assimilation are also analyzed,and the future research priorities and trends are discussed at the end of this paper.All of these can provide a theoretical basis for the further study of particle filter,so as to make a better use of it in data assimilation.
The sensor calibration accuracy is an important measurement of SAR satellite’s performance and data quality.Taking TerraSAR\|X for example,this paper studied the procedure and critical techniques of space\|borne SAR field calibration,including corner reflectors’ design and manufacture,corner reflectors’ arrangement in the field,corner reflectors’ status parameters calculation and the surveying scheme.Based on the field data and SAR image acquired synchronizedly,TerraSAR\|X’s geometric and radiometric calibration parameters were validated.The results demonstrated the effectiveness and accuracy of the field calibration procedure and critical techniques.
The Extend Fourier Amplitude Sensitivity Test (EFAST) was used to analyze the global sensitivity of SEBS model parameters in Yingke oasis station located in Heihe river basin from three points below:(1) the sensitivity of input parameters was analyzed,when sensible heat flux (H),latent heat flux (λE) and evaporation fraction (fr),respectively,are as the output parameter of SEBS model.(2) The sensitivity of surface parameters was analyzed,when the meteorology data was fixed.(3) Impact of parameters ranges on sensitivity index was analyzed by tuning input parameter ranges.The results showed that the model is sensitive to air temperature,wind speed,surface temperature and vegetation parameters.The impact of parameter interactions is slight for H and λE,but great for fr.When the meteorological data is fixed,the output variables are highly sensitive to surface temperature with high main sensitivity index (about 0.6).The sensitivity of input parameters is different using different parameter ranges.
With the ability to measure the whole of four Stokes parameters of sea surface radiation,the fully polarimetric microwave radiometer provides a new technique for measuring sea surface wind vector remote sensing.Because the amplitude of wind direction signal is only a few Kelvins,the absolute calibration accuracy of the third and fourth Stokes parameters brightness temperatures are needed at the tenth of Kelvin level,in order to adequately resolve their ambiguity on wind direction.based on the characteristics of wind direction harmonics,the calibration bias and instrument noise of WindSat polarimetric channels are analyzed by WindSat satellite measured data and collocated QuikSCAT oceanic products.The analysis results show that there are significant calibration biases in the WindSat 10.7,18.7 and 37.0 GHz frequency,and the biases are different on the different observation model,such as ascend orbit,descend orbit,fore swath and aft swath.In addition,the instrument noises of the third Stokes parameters of WindSat are less than 0.2 K,and the fourth Stokes parameters are less than 0.1 K.
The limb/occultation detection is an important way and direction for the satellite remote sensing.The inversion methods about the limb observation way could get a fine vertical distribution for the atmospheric elements.The most of the limb radiative transfer model have existed currently that are One\|dimensional model based on the horizontal homogeneous atmosphere.And they also ignore the impact of atmospheric polarization.The polarization is an important means for the atmospheric remote sensing,especially for the limb detection in the UV band.In order to meet the demands of limb detection in future,this study developed a limb model for solving vector radiative transfer in the horizontal inhomogeneous atmosphere based on successive order of scattering .And the impact of the horizontal inhomogeneous distribution has been considered in this model.The simulating value of the radiation intensity between research model and the LGLTRAN model that were developed by the Guo Xia et al.have been compared in the condition of exactly same environmental parameters.The result could be proved the reasonableness of this model.At the same time,we also made some simulation and simple analysis for the Stokes vector in the case of different atmospheric stratification.
Forest classification plays an important role in understanding the structure and function of forest ecosystem.Because the type of topography and forest richness were complex,the Landsat TM images was divided into 16 corresponding pieces according to administrative regionalization in Yunnan province.Firstly,based on 18 variations,such as TM 1~5 and 7,and vegetation index,KT transform results,Principal Component(PC) transform results,DEM data and so on,this study computed the mean variation of spectral values of training sets and the relationship between the spectral values and its frequency in the paper.Secondly,using intersection point algorithm,the thresholds of each classification were obtained and decision tree was established to separate each class step by step.Thirdly,by merging the classification results of 16 pieces into one large area,the total classification result of Yunnan forest was achieved.Finally,we selected Shangri\|la County as a validation case,and the results derived by decision tree method,which were compared with those of maximum likelihood approach.It shows that the total accuracy of the former is 86.61% (Kappa=0.80),while the latter is 74.39% (Kappa=0.63).The result indicates that the decision tree method is suitable and effective for forest classification and can provide basic data support for the study of forest LAI and biomass estimation in Yunnan province.
As the first spaceborne microwave remote sensing instrument in China,the Microwave Radiation Imager on FY\|3B has aroused widespread concern from science researchers in some fields,such as data quality and applications.In order to fully understand the performance of MWRI data,method comparison between satellites is used in this paper.The polar region ice sheet is as the study area,and AMSR\|E data is considered as the standard data.Data Processing and accessment are implemented on MWRI brightness temperature data.It is indicated that there is no significantly difference between MWRI brightness temperature data and AMSR\|E.The brightness temperature varies with target data,and there is a positive correlation between the both.Compared with different channels,the results of 10 v and 36 v are worst with more than 3 K absolute deviation and root mean square error.The result of 10,18 and 23 h are the best,with mean deviation less than 0.8 K,and root mean square error is less than 1.4 K and there is more than 98% bias is less than 3 K .The results of linear regression analysis show that there is obvious consistency in data trend.The polarization of H is better than V.The resules of different observation intervals illustrate that the longer the time interval is,the bigger V polarization contrast differences are.
Hyperspectral image has increasingly wide applications in remote sensing field.However,its own high dimension data and high redundant in inter\|band takes certain difficulties.For this issue,the paper put forward a novel algorithm by improving the firefly algorithm based on between\|class separability criteria to precede band selection.Specifically,motivated by firefly algorithm,the idea and framework using bio\|inspired algorithm for hyperspectral band selection are described,similarity matrix in inter\|band is designed,Euclidean distance,J\|M distance,spectral information divergence as between\|class separability criterion are used for objective function,and the discriminant bands based on the merits of target value are chosen.In addition,the experiments and performance assessment were conducted by HYDICE Washington DC Mall and HyMap Purdue Campus data.The experiment results have proved the promising ability of the proposed method for hyperspectral band selection.
This paper analyses the Line Mixing(LM) and Collision\|Induced Absorption(CIA)effect on the absorption in the 1.61,2.06 and 0.76 μm regions which used in the TanSat for retrieval of CO2.The LM effect on the cross section in 1.61 and 2.06 μm is about 2%.At the same time,analysing the LM and CIA effect on the absorption cross section in the 0.76 μm of O2.The relative effect of LM and CIA can reach 75% in the wings and about 2% at the P and R branch peaks.LM narrows the profile,leads to dips in the wings and to peaks in the centers to reduce the structural error by Voigt profile.This study analyses the change of absorption cross section and LM effect according to pressure and temperature,the LM effect is more significant with the greater pressure and the lower temperature.In the three bands,we chose the reasonable pressure,temperature and wavenumber interval to build the absorption cross section lookup table in advance to satisfy retrieval data throughput.
Light\|absorbing particulate pollutants in snow decrease snow reflectance significantly bring important effects on energy balance and climate change.The reduction of snow reflectance due to pollutants is an important and active research topic.In this study,a controlled experiment was carried out to measure the spectral reflectance of snow,in which the different synthetic pollutant amount was added.The study area is located in Fuyun county of Xinjiang Province where the seasonal snow area is representative of arid and semiarid region,and the spectral reflectance of the snow pollutant used in our experiment was also measured.The impacts of the pollutant with different concentrations on snow reflectance were analyzed quantitatively by building a linear mixing model using measurements.The results show that the relationship between reduction of snow reflectance and the pollutant concentration was nonlinear.As total pollutant concentration increased,the impact of 1 ppm(parts per million)pollutant on snow reflectance decreased.In experiment,the impact decreased by more than one\|third when the concentration increased from 1 813 ppm to 9 507 ppm.Moreover,this study found that the physical characteristics(such as grain size and shape)of snow and pollutants also played important role in the variation of dirty snow reflectance besides of the pollutant concentration.
Debris flow is one of the serious geologic hazards in southwest of China."5.12" Wenchuan earthquake constitute a large number of hidden unstable slopes.In heavy rain conditions,these potentially unstable slopes are very easy to form sudden mudslides.For this purpose,on the basis of the watershed division,the Wenchuan County is divided into different watershed unit using different scales,select the evaluation factor,evaluation model,and achieve the potential sudden whole Wenchuan County Debris Flow Hazard Assessment.The assessment results can provide reference and basis for Wenchuan debris flow disaster prevention and mitigation planning.Meanwhile,the results show that if the watershed scale was 200 000,300 000,400 000,the high and above risk of potential sudden debris flow were 58,59,30,accounting for 43.61%,67.05%,57.59% percent of the total watershed unit.
Considering the correlation between channels of hyper\|spectral atmospheric infrared sounder,as well as the timeliness of variational assimilation,channel selection is necessary.The principal component\|stepwise regression method is adopted for AIRS channel selection in this paper.Due to the fact that CO2 short wave channel easily affected by sunlight during the day,this study consider the channel combination during daytime and nighttime respectively.In the process of specific implementation:First,the channels are pre\|processed.Then,the principal components of Jacobi matrix about the temperature and humidity are analyzed respectively and some principal components with great effects are selected.By using the stepwise regression method,the channels with great affect on principal components are selected to form a channel subset.Further,according to the experience and actual observation data,principal components dual\|zone regression method is adopted for channel selection,in order to make an achievement of optimizing the global and take the local into account,with the idea of partition.Research results indicate that:①it is very necessary to rationally choose the channels when retrieves atmospheric temperature and humidity with AIRS data;② The channel combination obtains from principal components dual\|zone regression,and retrieving atmospheric temperature and humidity error are smaller than iterative method based on information entropy as a whole.
Due to the climate change and human activities,the impact of desertification in Qinghai becomes more serious in recent decades.Fast dynamic monitoring desertification by an effective technology is the necessary means to combating desertification.Based on the dimidiate pixel model,the vegetation coverage was estimated for Qinghai province using the MODIS data product—Enhanced Vegetation Index (EVI) from 2000 to 2012.The land desertification status and dynamic changes were rapidly monitored and assessed using vegetation coverage as the main index.The results show that remote sensing area of desertification in Qinghai is basically same as the conclusions of “A Bulletin of Status Quo of Desertification and Sandification in China”.The spatial distribution characteristics of land desertification in Qinghai did not change obviously during the 13 years.However,the severe desertification area shows obviously downward trend.The sandification and desertification status in Qinghai are improved,and the ecological communities productivity and vegetation coverage of grassland in some regions gradually increase.This indicates that recent years forestry engineering and desertification control measures implemented in Qinghai gradually which exhibit ecological benefits and play a practical role.
Island has high biodiversity and important ecological function.In recent years,China’s ecological environment of the islands are suffering from serious damage by human activities.Nowadays,we need investigate the island features to provide technical support for the island ecological environment protection.By using high spatial resolution images,based on coral reefs island remote sensing properties,with the aid of the Support Vector Machine(SVM)which has good classification performance,especially with small sample,nonlinear and high dimensional feature space,has good generalization ability to build structural support vector machine(SVM)theory and algorithm.According to the coral information extraction,comparing the kernel function of SVM to obtain ideal function,image preprocessing can improve the classification accuracy.
Using the SST data issued by NCEP/NCAR from 1950 to 1979,We have obtained three phases of ENSO envents,which are called as cold phase,normal phase,and warm phase.For each phase,the NDVI average values and the pictures of standardized departure are corresponding to growing season(Apr.\|Sep.)and winter season (Oct.\|Dec.)have been extracted from the NOAA\|AVHRR NDVI images with 1 km resolution ratio in 1982~2005.The results show that vegetation conditions of cold phase are better than the other two phases in growing season,however,in winter season,the vegetation conditions are the best in normal phase,followed by in worm phase,and the worst are in cold phase.Furthermore,the results also present obvious difference in the vegetation change between north and south of Tibetan Plateau.
In recent years,land degradation characterized by rocky desertification in karst areas of southwestern China becomes one of China’s three major ecological problems.Karst areas are areas of high complexity surface,and with high spatial and temporal heterogeneity and seriously mixed pixels.Vegetation,bare rock,bare soil are typical features of Karst regions.Thus it makes it more difficult to extract the key indicators like the fractional cover of vegetation and exposed bedrock which is required to evaluate how serious the rocky desertification is.Hyperspectral remote sensing has unique advantages on unmixing,and can get the abundance of surface features endmember.First,ground tests showed the spectral indices could characterize the feature coverage.Secondly,based on the data of Hyperion hyperspectral images,the study proposed endmember extraction of three types of classes feature from Hyperspectral Images,and estimate abundance by semi\|constrained and fully constrained linear spectral decomposition method.Results showed that:semi\|constrained linear decomposition method was better than fully constrained,and its abundance of inversed vegetation,bare soil,bare rock was much more suited for spectral indices.The proposed feature abundance was able to characterize its coverage with R20.92,0.66 and 0.84,respectively.The linear unmixing method inversed the feature abundance to extract the Karst indicator was feasible.This study indicates that hyperspectral remote sensing laid the foundation for the karst rocky desertification assessment and monitoring.
A method,means of stratified non\|homogeneous surface,is used to design the sensor network in the middle reach of Heihe river basin to capture the surface freeze/thaw cycles.This method decomposes a heterogeneous surface into several subareas which are homogeneous.Its estimation variance is calculated by the kriging technique and stratified sampling.With the goal of minimizing the estimation variance,nodes of the sensor network are optimally distributed in the study area by the simulated annealing algorithm.The result shows that optimal samples have good representativeness,and high density of samples are assigned to the stronger heterogeneous regions.
Using the Mean Cloud Amount (MCA) data of 1984~2009 annual International Satellite Cloud Climatology Project (ISCCP) D2 data set,time series features of mean cloud amount in different regions of Central Asia are extracted by time series analysis method.Using the time series stationary features of mean cloud amount,cloud amount in Central Asia is divided into two categories:the monthly mean cloud amount time series stationary type and non\|stationary type,and then the non\|stationary type is divided into two categories:the monthly mean cloud amount time series ordinary non\|stationary type and seasonal non\|stationary type by time series seasonal features of mean cloud amount.At last,mean cloud amount in Central Asia is divided into three categories:the monthly mean cloud amount time series stationary type,namely,stationary type,the monthly mean cloud amount time series ordinary non\|stationary type,namely,ordinary non\|stationary type,and the monthly mean cloud amount time series seasonal non\|stationary type,namely,seasonal non\|stationary type.The classification results show that the seasonal non\|stationary type accounts for 62.37% of the study area,the ordinary non\|stationary type and stationary type respectively account for 31.18% and 6.45% of the study area respectly.The overall cloud amount in Central Asia shows non\|stationary characteristics.The seasonal non\|stationary type areas are mainly concentrated in Kazakhstan and Xinjiang region,the ordinary non\|stationary type areas are mainly in the northern and central east Kazakhstan extending to a strip region of Xinjiang and the stationary type areas are mainly in Amu Darya River Basin.The mean cloud amount results show that,the overall mean cloud amount in Central Asia is 64.69% from 1984 to 2009 with a maximum of 66.50% in 1984 and a minimum of 62.38% in 1999.The difference between maximum and minimum mean cloud amount is 4.12%.The mean cloud amount in Central Asia has a downward trend at first,and then has upward trend.There was a shock in the middle,but the overall shows a downward trend.
It’s difficult to directly measure or calculate the surface area.The surface area is often replaced by projected area.In the plain area,the differences between these two kinds of area are not worth being noticed,but in the mountainous area,the differences are obvious.In order to calculate the surface area of Hengduan mountainous area which has the dramatic mountain terrain ,Qamdo County in Tibet was chose as the study area.The geology significance of the surface roughness is worked out,and based on that the surface area has been calculated.The differences between the surface area and projected area have been quantitatively analyzed.The results showed that compared with the same land use,the differences of the higher the solution of DEM data is the more obvious between the surface area and projected area.The differences of the two kinds of area for grassland,shrub land,forest land,open forest land and wasteland were obviously,which were accounted for 11.275%,13.447%,16.586%,13.644% and 17.282% of their projected area.The differences between the surface area and projected area in the dramatic mountain terrain can’t be ignored.The differences of the two kinds of area increased with the increasing of slope,especially when the slope is more than 75 degree.Based on the geology significance of the surface roughness,the landscape surface area which can offer more scientific and effective data for land use change detection,forest resources investigation,crop production estimation with establishing the surface roughness database in typical area can be quickly access to.
In the field of earth science,Self\|similarity is very common.The wall rock alteration is reflected on the image as a geo\|science character,whose pixel brightness and area model conform to the fractal theory.Using the summation method of fractal theory for the extraction of remote sensing alteration anomalies and classification can make up for the threshold method and gray level histogram visual method to determine the mutation point.Finally obtaining the fractal features of remote sensing alteration anomaly in the study area by fractal box dimension method.This paper taks ETM+ remote sensing image data of Maoping lead\|zinc deposit as the information source,which is used for Principal Component Analysis (PCA) reflect the mining area of remote sensing alteration anomaly of principal components,and statistics the principal component of pixel grey value and the number of pixels.Then fractal method is used to analyze alteration anomaly information and divide the alteration abnormal levels.Finally using fractal box dimension method get Maoping lead\|zinc deposit fractal features of remote sensing alteration anomaly information and compared with the traditional threshold method.The study show that the remote sensing alteration information of the study area has characteristics of statistical fractal and the fractal method to determine abnormal limit is more accurate than the traditional methods of threshold selection alteration threshold;There is a positive correlation between fractal dimension values and the abnormal level;Fractal dimension of high value areas and the alteration information concentration areas is consistent.
According to the needs of the project this study can provide reliable reference for the optimization and widely use of the data of ZY\|3 based on the scientific and applied researches on the images features of ZY\|3.In this paper,spectral characters in ZY\|3 are analyzed from three aspects: single band data information content,multiband data correlation and the class separability of each band data.Through the analysis,we can get that Band4 of ZY\|3 is the most satisfying band for its richest information content,minimal band correlation and maximum spectral range between classes.The following are band3 and band2,and band1 is the least satisfying band in all indexes.The further analysis of Resolution merge of Multispectral and Panchromatic data after the combination of band2 & band3 shows the images of principal component transformation have the best visual effect,while the images of Brovey Transform have the richest information content.
The research estimated the vegetation coverage using dimidiate pixel model based on three TM images in 2010 in Beijing.And the paper shows the seasonal changes of vegetation coverage based on the real vegetation coverage in the field.The paper respectively calculates the NDVIsoil and NDVIveg which are the parameters of dimidiate pixel model in three periods,and gets vegetation coverage in Beijing.In addition,this study measured the real vegetation coverage in the field and got the seasonal change curves of vegetation coverage.The relationship between real values and estimated values of vegetation coverage is significant.The data nearly aggregated around the line 1∶1 in the plot of real values and estimated values.The correlation coefficient between them are all over 0.8 during the three periods.They are closely related to each other.The curves of seasonal change getting from TM image are similar to those of the real vegetation coverage during the three periods in 2010.
For the current automatic image registration based on SIFT,feature point matching algorithm is time\|consuming,in addition,the changes of multi\|temporal images affect the accuracy of registration,this paper proposes a SIFT\|based feature of the “low\|resolution matching,high resolution authentication” hierarchical image registration algorithm to improve the above issues.In the proposed algorithm,affine transformation model is established in low\|resolution pyramid images and sequentially evaluated and revised by match points in high resolution pyramid images.Experimental results show that the improved SIFT algorithm can reduce the time complexity with rather considerable accuracy.
Metadata standard is essential for regularizing data resources management and maximizing data sharing.It also is capable of regulating and normalizing data entering process to databases,to ensure data integrity and quality.Metadata standard should be constructed fully aware of disciplinary characteristics that the data reflect.Based on that,a variety of databases and data information systems can be built to facilitate data sharing.It was intended to be developed a metadata standard for permafrost investigation data that captures principal characteristics of permafrost and relevant data.It uses ISO 19115,the international metadata standard of spatial data,as the primary template and references from other country standards including ecological,environmental,meteorological,and geological standards.After a simple introduction of the compilation principals,metadata entities composed of this standard and its hierarchical relationship between entities,as well as its major features,are presented in details.The extensions and modifications from ISO 19115 are also described.Finally,The study established upon this metadata standard,an application,titled "Information System for the Permafrost Background Investigation over the Qinghai\|Tibet Plateau" was built.Through this work,permafrost data management and sharing are expected to be promoted to a higher level.
The appearance of LiDAR technology provides a new method for automatic vehicle detection.In order to detect vehicle object from LiDAR data,according to the property features of different objects,a new method of automatic urban vehicle detection from airborne LiDAR data with aerial image is proposed in this paper.Firstly,it is classified ground and non\|ground points using LiDAR filtering with morphological opening by reconstruction.Secondly,with the help of aerial image and its Normalized Differential Vegetation Index (NDVI) feature,it could classify LiDAR non\|ground points into vegetation and non\|vegetation objects.Finally,On the basis of non\|vegetation objects,it could separate vehicle objects automatically from other non\|vehicle objects by shape feature and height property.Three regions has been used to verify the feasibility and reliability of this method.The experiment results show that the proposed method can effectively extract vehicle objects.The mean of correctness and completeness of this method can reach 95% and 85% respectively,which can meet the practical requirements.
It is a new direction to expand the Geographical Information System (GIS) application areas by integrating GIS technologies with Bibliometrics.In recent 10 years,GIS was applied in bibliometrics more and more extensively.A bibliometric GIS tool was developed based on the open source GIS software DotSpatial,which is aim to help the non\|GIS professionals to make the operations of spatial information mining,query,map\|making and analysis.Its main functions include literature data import,geo\|spatial linked information extraction,indicator calculation,layer operation,view operation,and map output.The analysis results show that this tool can realize the spatial display and map\|making of the scientific literature.Therefore,it is a practical thematic GIS tool.
